The Agent Skills Directory

Category 1: Prompt Injection (SAFE): The skill contains purely instructional content for debugging and does not attempt to override AI safety constraints or hijack agent behavior.
Category 2: Data Exposure & Exfiltration (SAFE): No sensitive file paths, hardcoded credentials, or unauthorized network communication patterns were identified. Code examples use placeholder data (e.g., 'Alice', 'user_5432') and common testing endpoints like 'localhost'.
Category 3: Obfuscation (SAFE): All content is provided in clear-text markdown. No Base64 encoding, zero-width characters, or homoglyphs were detected.
Category 4: Unverifiable Dependencies & Remote Code Execution (SAFE): The skill references standard developer tools (git, pnpm, curl) for local environment testing and links to reputable academic sources (MIT, Cornell). It does not execute remote scripts or download untrusted packages.
Category 5: Privilege Escalation (SAFE): No commands for privilege escalation (e.g., sudo, chmod 777) or unauthorized system modifications were found.
Category 6: Persistence Mechanisms (SAFE): The skill does not attempt to modify shell profiles, startup services, or scheduled tasks.
Category 7: Metadata Poisoning (SAFE): Metadata fields are accurate and consistent with the skill's stated purpose of providing debugging best practices.
Category 8: Indirect Prompt Injection (LOW): As an instructional skill for coding agents, it defines behavior for analyzing code. While it influences agent reasoning, it does not process untrusted external data in a way that creates a vulnerability surface.
Category 9: Time-Delayed / Conditional Attacks (SAFE): No logic was found that triggers behavior based on dates, environment variables, or other external conditions.
Category 10: Dynamic Execution (SAFE): The skill provides static code examples for educational comparison and does not perform unsafe runtime evaluation or deserialization of data.

debug